A longitudinal study is proposed to identify predictors of discord and instability over the first four years of marriage. The long-term goal of this research is to understand the determinants of marital discord and dissolution in early marriage so that effective prevention programs can be developed. The model guiding this research proposes that marital quality and stability are determined by (a) how individuals and couples contend with and resolve difficulties and transitions, (b) the demographic, personality, and experiential factors that spouses bring to marriage (e.g., experiences in their families of origin), and (c) the stressful events that spouses and couples encounter. To test this model, 200 couples will (a) discuss marital difficulties in a laboratory session shortly after marriage and again one year later, and also complete self-report measures of how they cope with difficulties; (b) complete measures of personality and demography in the initial laboratory visit, and also participate in structured interviews about their family of origin and premarital relationships; and (c) report on the stressful events and circumstances in their lives, on questionnaires administered during the laboratory visits and via mail at 6-month intervals over the four years of data collection. In addition, spouses will be asked to complete brief measures of marital quality and stability in the laboratory visits and via mail at 6-month intervals throughout the project. Interactions will be coded by trained observers using an established observational coding system. Hypotheses derived from the model will be tested using structural equation modeling and survival analysis. In short, the proposed study uses observational, self-report, and interview data collected over the first four years of marriage to delineate the factors that increase risk of discord, separation, and divorce. With this information, empirically-based prevention programs for marital dysfunction can be designed.